The proposed five-year training program is designed to provide the principal investigator with the background in microarray-based genome-wide epigenomic profiling directed towards translational brain cancer research. As an Assistant Professor, the PI is a clinically active neuro-oncologist with designated laboratory space. This training program will enable him to enter a new area of research that is directly synergistic with his clinical efforts as a co-principal investigator on a recently opened UCLA/Genentech clinical trial (target enrollment of 70 patients) evaluating the use of bevacizumab (Avastin) in combination with temozolomide and radiation for patients with newly-diagnosed glioblastoma. Critical determinants in his training program are the mentor, Dr. Stan Nelson, access to the UCLA Neuro-oncology database, and active support of the UCLA brain tumor research environment. Dr. Nelson is a recognized leader in the field of genome-wide array based analysis and has been active in large-scale genetic characterization of patient glioblastoma tissue in the UCLA Neuro-oncology database, a large clinically annotated collection of patient tissue, cell lines, and radiographic images. The potential for powerful correlative studies harnessing this database have been realized by a large number of recent studies from the UCLA group. The overall goal of the research plan is based on the recent discovery that glioblastoma patients whose tumors have O(6)-Methylguanine DNA methyltransferase (MGMT) promoter methylation demonstrate better survival and response to temozolomide (TMZ). This finding emphasizes the importance of aberrant methylation in brain cancer. We hypothesize that genome-wide methylation profiling of glioma will lead to the discovery of novel methylated genes and signatures of coordinately methylated genes, and we demonstrate the feasibility of using differential methylation hybridization (DMH) to achieve these goals. The specific aims are: 1) to understand the role of promoter methylation in astrocytoma transformation by comparing genome-wide methylation profiles of low grade astrocytomas, anaplastic astrocytomas, primary and secondary glioblastomas, and normal brain samples, 2) to determine genome-wide methylation, gene expression, and loss of heterozygosity (LOH) profiles associated with response to a novel upfront therapy combining bevacizumab with radiation therapy and TMZ, 3) to examine the role of DNA methyltransferase 1 (DNMT1) in glioblastoma methylation by profiling genome-wide methylation changes in glioblastoma cell lines resulting from pharmacologic or genetic modulation of DNMT1 activity. The overall goal of this proposal is to improve therapeutic outcomes for glioma patients by developing personalized approaches based on prospective molecular characterization of tumor samples.